Every studio that runs ads on two channels gets the same monthly report. One channel delivered leads at $35 each. The other delivered them at $75. The obvious move is to shift the budget to the cheap one, and the agency usually suggests it.
It is the obvious move, and it is often backwards. A lead is not what you are buying. You are buying a member, and the two channels turn leads into members at very different rates for a reason that has nothing to do with skill. This piece follows one month of leads from both channels to membership at one invented studio, Alturra Studios, with 240 members at $169 a month. It is one invented month with small counts. Your own tally may come out the other way, which is exactly why you need one.
The comparison in one minute
- The two channels do different jobs. Search reaches people who are already looking. Social reaches people who were not. Expect different lead prices and different conversion.
- Compare cost per member, never cost per lead. That needs your own record of how each member heard about you, because no ad platform can see who joined.
- Follow-up changes the answer more than the platform does. A lead who was not looking cools in hours. Reach her the same day and the arithmetic moves.
| February's leads, followed | Meta | What it says | |
|---|---|---|---|
| Spend | $1,995 | $1,500 | what the month cost |
| Leads | 57 | 20 | hands raised |
| Cost per lead | $35 | $75 | the number the report leads with |
| Lead → intro start | 6 of 57 = 10.5% | 6 of 20 = 30.0% | how many were really looking |
| Intro → member | 2 of 6 = 33.3% | 4 of 6 = 66.7% | how serious the visitors were |
| Cost per member | $998 | $375 | the number that decides the budget |
| Payback at $169 dues | 5.9 months | 2.2 months | how long the studio funds each member |
Which is better for a fitness studio, Facebook ads or Google Ads?
Neither in general, because they are not doing the same job. Understanding the difference explains every number in the table above.
She typed "yoga near me" and your ad answered. The intent was hers before you spent anything.
- Fewer leads, each closer to booking
- Limited by how many people search nearby
- Send her to: the class schedule and the intro offer
She was scrolling and your ad interrupted her. The interest is real and it is minutes old.
- More leads, most of whom never book
- Limited by how often the same people see the ad
- Send her to: one class, one offer, one button
Neither job is better. A studio that only buys search is fishing in a small pond: it catches most of the people who are looking and none of the people who would have loved the place had they heard of it. A studio that only buys social is paying to introduce itself to strangers and then, too often, not following up. There is a case for some of each. The question is how much, and that is where the tally comes in.
Before either, do the free version of search. People who look for a studio by name or by neighborhood see your Google listing before any ad, and keeping that listing current costs nothing. The local search piece covers the rest of the unpaid side.
Why is cost per lead the wrong way to compare them?
Because the two channels mean different things by "lead". A search lead is someone who went looking. A social lead is someone who tapped a button. Counting them as the same unit and dividing is the mistake.
Follow February's leads forward and the difference appears at every step. Alturra's signup form asks every new account how they heard about the studio, so each lead can be traced.
This is the same February cohort that the funnel piece follows as a whole: 90 leads, 15 intro starts, 9 members. Splitting it by source shows that the average was hiding two very different channels. Now divide each channel's spending by its members.
Two members and four members are small numbers, and one person either way changes the picture. Do not move a budget on one month. Keep the tally for a quarter, then read it. What the month does show reliably is the shape: cheap leads, expensive members on one side, the reverse on the other. The acquisition cost piece explains why the blended figure, $233 that month, could not have told you any of this.
Why are Facebook leads cheap but hard to convert?
Because the interest is minutes old, and it fades in hours. Someone who searched has a need that will still be there on Thursday. Someone who tapped an ad between two videos has a passing thought. That is not a flaw in the channel. It is what the channel is, and it sets the terms for using it well.
The terms are speed and a real step. Of Alturra's 57 Meta leads, 51 never started an intro. The desk reached them when it had time, which for many was two or three days later, by which point the thought had passed. Suppose instead every one had a call or a message from a named person the same day.
That is the most useful finding in this piece. Before deciding that a channel does not work, check how its leads were treated. The speed-to-lead piece makes the case for the same-day call, and the five follow-up messages give the desk the words. The other half is the ad itself: ask for a step that costs the lead something, buying the intro offer or booking a specific class, and fewer people will raise a hand they did not mean.
How should you split the budget between them?
Fund the channel with the cheaper members first, up to its ceiling, then the other. Each channel has a ceiling, and they are different kinds.
- Search runs out of searchers. Only so many people in your area look for a studio in a month. Once your ad reaches most of them, extra budget buys the same clicks at a higher price. The sign is cost per lead climbing while leads stay flat.
- Social runs out of fresh eyes. The same people see the same ad until they stop noticing it. The sign is cost per lead climbing while nothing else changed. New pictures and a new offer buy more room than more money does.
For Alturra the tally says: grow Google until its cost per lead starts to rise, and hold Meta where it is until same-day follow-up is working, then read it again. That is a decision about the next $500, not a verdict on either channel. The marketing budget piece sets the total; this is only the split.
How do you run a clean test?
Change one thing, run it a full month, and judge it at day 90. Most ad tests fail before they start, because three things changed at once and the result was read after a week.
The wait is the hard part. A month of leads takes about 90 days to finish becoming members when the intro offer runs 30, which is the argument in the intro conversion piece for reading closed cohorts. While you wait, cost per lead and lead to intro are the early readings. They can tell you a channel is broken. They cannot tell you it works.
What to do this week:
- Make "how did you hear about us?" a required field at signup, with five or six fixed answers.
- Check that both channels advertise the same offer and land on a page where it can be bought in one step.
- Pick a class with room and name it in the ad. Do not advertise the hours that already waitlist.
- Time your follow-up for a week: how long between a lead arriving and a person contacting her?
- Put a date in the calendar 90 days after month-end to read the first tally.
What can a dashboard tell you, and what can it not?
It can show spend, leads and cost per lead for each channel. It cannot honestly split members by channel. That limit is real, and a tool that claims to cross it is guessing.
The two ad platforms count different events, so their costs sit side by side and are never added into one "cost per lead". And each platform's own count of purchases is what its tracking saw, which is not the same as new members in your booking platform. The bridge between a channel and a member is the answer a person typed at signup. Keep that record and you have something no report can give you.
Where this lives
Half of this comparison can be kept current for you, and the other half is yours. Xyzios works with Mindbody and is an approved Mariana Tek integration, and it connects to Google Ads and Meta. It shows spend, leads and cost per lead for each channel, side by side and never blended, and one cost per new member that is not split by channel. It watches both accounts and emails the owner when a channel goes dark, spend surges, cost per lead spikes or the leads stop, so a broken campaign does not spend quietly for two weeks. The integrations page lists what connects.
It also works the part that moved the answer most. Each morning the Sales Desk lists the new leads and intro-offer holders worth a call, in the order they go cold, for a person to work. Nothing is sent from that list. How lead follow-up works covers it. Xyzios does not draft a lead's first message; the call is the studio's.
Start the tally this week. The math of one member shows what a $375 or a $998 member is worth over her stay, the leak calculator shows what the members you already have are worth keeping, and the two-minute studio check will show where your studio should look first.